Recently direct evidence supporting the contribution of effector cells derived from the donor to the anti-leukemic effects of a marrow graft has been provided by several groups including our own, who have found that among patients transplanted for CML over 70 percent of recipients of unmodified or T cell depleted marrow allografts who have relapsed post transplant can be reinduced into durable molecular remissions by treatment with infusions of high doses of donor-derived peripheral blood monounclear cells (PBMC). At present, the nature of the leukemia reactive cells(s) in adoptively transferred PBMC that incude remissions is unknown. Such leukemia-selective effectors may include T-cells reactive against leukemia-specific antigens or alloantigens preferentially expressed on leukemic cells or reactive against leukemia-specific oncogene fusion products, as recently described by our group, as well as natural killer cells exhibiting HLA-non-restricted cytotoxic activity. This project will focus on the development of strategies of adoptive cell therapy, and T cell therapy generated in vivo (active specific immunotherapy). incorporating selected or targeted effector cells that are able to distinguish leukemic cells. Human leukemias bearing fusion gene tragets that will serve as models in this project include CML and Ph+ ALL (p210 and p190 targets), AML-M2, t(8-21) (p92) and AML-M4, inv (16) (p67). The Aims are: 1) To develop effective strategies for generating and selecting potentially immunogenic leukemia fusion gene peptides. 2) To sensitize and propagate in vitro, leukemia associated fusion gene peptide-specific T lymphocytes and to characterize these cells as to their antigen-specificity and their relative proliferative and cytotoxic responses against HLA-matched peptide loaded normal targets in comparison with leukemic cells bearing the fusion gene. 3) To assess the leukemia-specific activity of fusion gene peptide-specific T cells, in comparison with alloreactive T cells and NK cells both in vitro and in vivo, after adoptive transfer into xenografted SCID mice bearing the targeted human leukemia or a normal hematopoietic cell graft. These experiments constitute a "proof- of-principle" for this approach. The results of these preclinical studies will be used to formulate clinical trials to assess the antileukemic activity of vaccinating donors, or patients with CML, PH+ALL or AMLs bearing fusion gene products.